The underestimation of spring precipitation over South China is caused by the weak simulations of the dynamic motion in CMIP6 models

This study investigates the performance of the 23 CMIP6 models in simulating spring precipitation over China and preliminarily explores the reasons for the underestimation of spring precipitation over South China (SC). The results show that the CMIP6 models can generally capture spring precipitation over China but have distinct underestimations in SC. In addition to analysing the results of the multimodel ensemble mean (MME), based on the regional mean spring precipitation differences over SC, we also analyse five models with greater overestimated precipitation amounts composing the high-value MME (H_MME) and five models with greater underestimated precipitation amounts composing the low-value MME (L_MME) as comparisons. Moisture budget analysis reveals that the underestimated dynamic component of vertical moisture advection mainly causes the negative biases of spring precipitation over SC. Further analyses indicate that the underestimation of the intensity of the East Asian subtropical jet (EASJ) generally causes weakened secondary meridional circulation to the south of 30 degrees N and moisture convergence over SC in CMIP6 models. Consequently, less moistened air would be lifted to high levels, providing an unfavourable background for the occurrence of spring precipitation and inducing underestimation compared with the observations. The modelling EASJ shows much stronger (weaker) H_MME (L_MME), which generally enhances (weakens) the dynamic component of the secondary circulation over SC, favourable for the increase (decrease) in spring precipitation.